Single sideband modulation is basically double-sideband modulation
with one of the sidebands removed. This means that a single sideband
signal takes less bandwidth than a double sideband signal; all of the
information in the message signal is contained within both of the
sidebands, so it is completely acceptable to eliminate
one. Practically, single sideband modulation is harder to implement
than either DSB or AM. It is also more difficult to understand, and the math
is more difficult. Its big advantage is the bandwidth savings it
provides, and the fact that it is much more efficent than AM. SSB modulation has a 9 dB signal to noise advantage over AM, meaning that at -6 dB an AM signal becomes unusable, but a SSB signal can still get through.
Single sideband modulation can be implemented in at least two
ways. The first way would be to multiply the message signal by a
carrier, which gives you a DSB signal. A filter can then be applied to
remove one of the sidebands. Practically, it is very difficult to
implement a filter with a sharp cutoff, so distortion results. If this
method is used, the filter is usually built using crystal filter
technology. Mathematically, this modulation of a message signal can be
express by using the Hilbert Transform.
Phase-shift modulation can also be used to produce a single
sideband signal. This works by taking the two copied of the message
signal, one of which is multiplied by a carrier cosine, while the
other is transformed by the Hilbert Transform before being
multiplied by a carrier sine. These two signals are then added to or
subtracted from each other to yield lower sideband or upper sideband
signals.
Several methods may be used to demodulate single sideband
signals. The first method is very similar to that used for DSB signals; just multiplication by a demodulation carrier
and lowpass filter the result. As for DSB, if the phases of the
modulation carrier and the demodulation carrier aren't aligned with
each other almost perfectly, serious distortion of the message signal
can result.
Another method of SSB demodulation is called carrier reinsertion.
This involves adding the output of an oscillator to the received
signal, and then feeding that signal into an envelope detector, the
output of which is
yd(t)~= (1/2)Acm(t) + K
where m(t) is the original message signal, and Ac and K
are constants. The message signal can then be extracted from the
envelope detector output with relative ease. Again, frequency and
phase of the demodulation carrier and the modulation carrier must
match up.
Credit goes to Rancid_Pickle for the signal-to-noise ratio information. Thanks!